first said something interesting. Recently my little girlfriend always thought my figure was too greasy, so I went into the weight loss stage. Finally, yesterday, 66.7% of all the people I met said I was miserable and skinny. I was so happy to go home and weigh it and tell me I’m still a fat man.
thought for a long time. I think the 66.7% of them lied to me: "the 2 men are so bad.". (yes, I met 3 people that day)
this story tells us: data base is too small, will affect data analysis
a lot of people have talked about the problem of conversion, but they are all aimed at one value. There are probably more ways to solve the problem than to think about it.
How many layers of screening did
users finally form the final valid customer?
what variables determine the customer’s willingness to divide himself into a particular category?
below is a brief summary of conversion rates at different levels, four modules, six levels,
upgrade any layer of conversion will make the final sales have improved, so when a road to the end, you can look at other ways. Each layer is optimized, and there will be unexpected results
traffic is king, but the most important thing we should do is to improve the quality of our own shop, so that people who go into the store stay in the store more and become effective users.
it makes me think of the evil some thin (allow me here God Tucao), they eat every day in all kinds of energy substances out of order, finally actually was a thin!! every day in front of you said: "I just eat all not fat ah what….. I really feel that they are not what good playfully!! their conversion rate so low waste much national food!
or talk about how to see the conversion rate of it.
a lot of times, if you can find something that users are interested in earlier than others, you’re the one who makes the most money.
so you should know what’s the best baby you can sell in a week,
to determine whether a baby may be a burst of terms:
has sent friends once confused, so many data software, why give the final conclusion is not the same,
that’s because their data models are based on different dimensions, so we need to know what their judgment is based on,
‘s selection of potential baby steps can be summed up as follows:
first step: look for potential babies in each category